DESCRIPTION: Adenoviral vectors are regarded as the most efficient gene transfer vehicles for in vivo human gene therapy. However, the currently available E1 E3-deleted vectors have had limited success because of relatively short-term expression in vivo. This is caused mainly by low-level leaky expression of viral genes retained in the vector, resulting in a CTL-mediated immune response by the host against the virus-transduced cells. Another cause of short-term expression is the lack of integration of adenoviral vector DNA into host chromosomes. Recently, a helper-dependent adenoviral vector system was generated by the PI to overcome the problem of CTL response. To obtain sustained expression from the vector in dividing cells, it is desirable to develop a strategy to maintain the vector DNA in the transduced cells, preferably by stable integration into the chromosome. In this proposal, approaches will be taken to design adenoviral vectors for stable integration into host chromosomes, and thus to obtain long-term expression from the vector. To achieve this goal, improved helper-dependent vectors will be developed to achieve higher vector titers with no contaminating helper virus for better in vivo use, and thus to overcome the current shortcomings of the helper-dependent adenoviral vectors. The integration efficiencies of adenoviral vector into host chromosomes will be analyzed in mice and in different cell lines to elucidate viral and cellular factors involved in chromosomal integration of adenoviral genomes. Ideally the transferred gene would be integrated into any target region on a chromosome to obtain tissue-specific and sustained expression. However, to date, no successful approaches for in vivo gene targeting have been reported. In the proposed work, a helper-dependent adenoviral vector with target chromosomal sequences will be constructed, and the efficiencies of homologous recombination will be evaluated both in vitro and in an in vivo mouse model.